The paper was accepted on the CV4MetAVER WORKSHOP ECVPR 2025.
With prompt development on Virtual Reality (VR) headsets, according to successful stereoscopic quality of experience it is important to bring more extra experiences. However, most of the metrics focus on the diversity of viewing experiences such as unpleasant views or image quality, and traditionally deal with data limits. Dealing with these posts, introduces the wide (closure of the content of the content), a new dataset containing the original sterereoscopic images and the performance of various different comments and art. The dataset is recorded with popular preferences collected in the VR headset, with a significant degree of consignment for user choice for all different headsettes. Additionally, we present an isted, new model of the stereo quality to test trained experience in our database. We show that the Istochee matches the popularity of people there are ways in which they compare Mono-stereo modeling methods.
- † work done while in apple
- Wezmann Institute of Science
- § Imassachusetts Institute of Technology
- ¶
